Navigation of Autonomous Mobile Robots in Diverse Terrain
Terrence P. Fries · 2020
Optimal path planning is essential for successful operation of mobile robots. Many applications require a robot to operate outdoors, frequently in diverse terrain conditions. These include earth and space exploration, surveillance, and search and rescue. Many existing path planning algorithms utilize either genetic algorithms (GAs) or fuzzy logic. However, these algorithms fail to consider the effect of terrain conditions when determining the optimal path. Additionally, they are unable to deliver real-time adjustments required for operation in a dynamic environment and focus solely on path planning without regard to trajectory planning required to facilitate movement. This paper proposes a navigation algorithm that utilizes a genetic algorithm to compute a near-optimal path through varying terrain and the associated trajectory planning required to follow that path. Fuzzy linguistic variables are used to represent the terrain conditions to address the uncertainty and imprecision inherent in terrain data. The navigation algorithm presented is robust, allowing rapid changes in the path to adapt to changes in dynamic environments.